Optimizing Drug Response Study Design in Patient-Derived Tumor Xenografts

Jessica Weiss; Nhu-An Pham; Melania Pintilie; Ming Li; Geoffrey Liu; Frances A Shepherd; Ming-Sound Tsao; Wei Xu

doi:10.1177/11769351221136056

Optimizing Drug Response Study Design in Patient-Derived Tumor Xenografts

Cancer Inform. 2022 Nov 22:21:11769351221136056. doi: 10.1177/11769351221136056. eCollection 2022.

Authors

Jessica Weiss¹, Nhu-An Pham², Melania Pintilie¹, Ming Li², Geoffrey Liu^{2

3

4}, Frances A Shepherd^{2

3}, Ming-Sound Tsao^{2

4

5}, Wei Xu^{1

6}

Affiliations

¹ Department of Biostatistics, Princess Margaret Cancer Centre, University Health Network, University of Toronto, Toronto, ON, Canada.
² Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada.
³ Department of Medicine, Division of Medical Oncology, University of Toronto, Toronto, ON, Canada.
⁴ Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada.
⁵ Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, Canada.
⁶ Department of Biostatistics, Dalla Lana School of Public Health, Toronto, ON, Canada.

Abstract

Patient-derived tumor xenograft (PDX) models were used to evaluate the effectiveness of preclinical anticancer agents. A design using 1 mouse per patient per drug (1 × 1 × 1) was considered practical for large-scale drug efficacy studies. We evaluated modifiable parameters that could increase the statistical power of this design based on our consolidated PDX experiments. Real studies were used as a reference to investigate the relationship between statistical power with treatment effect size, inter-mouse variation, and tumor measurement frequencies. Our results showed that large effect sizes could be detected at a significance level of .2 or .05 under a 1 × 1 × 1 design. We found that the minimum number of mice required to achieve 80% power at an alpha level of .05 under all situations explored was 21 mice per group for a small effect size and 5 mice per group for a medium effect size.

Keywords: 1 × 1 ×1; PDX; lung cancer; mixed-effect model; n = 1 experiment; replicates; therapy response; treatment effect size; tumor growth rate; xenograft.